Abstract
A detailed analysis of genotype-environment interactions was carried out among yields of six cultivars of hard red spring wheat grown at each of nine locations in five different years. Subdividing the sum of squares for genotype-environment interactions into components due to each cultivar indicated that the Finlay-Wilkinson method of measuring yield stability is of little value for wheat yield in western Canada. Conventional estimates of variance components due to the different types of genotype-environment interaction indicated that all except the genotype-year interaction were significant and important.